Getting started
Install one Benchmark and run one deterministic Evaluation.
Installation
You need Python 3.12 and uv 0.11.16.
git clone https://github.com/Linzwcs/EvoPolicyGym
cd EvoPolicyGym
uv sync --project environments/gymnasium/classic_control/cartpole --extra dev
uv run --project environments/gymnasium/classic_control/cartpole \
evopolicygym --version
Expected version output:
evopolicygym 0.3.0
This installs the EvoPolicyGym Kernel and the independent CartPole Benchmark.
Evaluate the baseline
Run the packaged baseline over five validation Episodes:
uv run --project environments/gymnasium/classic_control/cartpole python - <<'PY'
from cartpole import CartPoleBenchmark, baseline_program
from evopolicygym import EvaluationConfig, evaluate
from evopolicygym.execution import ProcessExecution
result = evaluate(
baseline_program(),
CartPoleBenchmark(),
execution=ProcessExecution.unsafe(),
config=EvaluationConfig(split="validation", episodes=5, seed=42),
)
print(result.feedback.score)
print(result.feedback.content)
PY
The result contains a scalar score, public Feedback, the Program digest, and sanitized Episode summaries.
Local process execution
ProcessExecution.unsafe() runs Policy code with your operating-system user
permissions. It is not a sandbox. Use it only with trusted code.